How Kanji Mastery is built

Learning approach

The app separates writing recall from reading recall because recognising a character, producing it from memory, and reading it inside a word are different skills. Tests update those abilities independently, while drills provide extra exposure without pretending that one successful guess means mastery.

Smart Review uses FSRS-6 scheduling. Instead of reviewing every character at a fixed interval, it estimates how difficult a card is, how stable the memory has become, and how likely you are to recall it now. Your Again, Hard, Good, and Easy responses adjust the next interval.

Data and editorial method

Character meanings and readings come from KANJIDIC2, vocabulary data comes from JMdict, and stroke-order diagrams come from KanjiVG. These projects are credited on every page and in the Terms of Service. Example sentences may be retrieved from Tatoeba.

Source data is organised into study activities, filtered by approximate JLPT level, and supplemented with learning workflows, progress tracking, confusion detection, and spaced repetition. Vocabulary examples are shown to put readings in context; a kanji keyword should be treated as a memory cue, not as a complete translation of every word containing that character.

About JLPT labels

The JLPT does not publish an official kanji syllabus. Levels on this site are community-sourced approximations used to organise study, not a claim that a character will or will not appear on an examination. Kanji Mastery focuses on kanji practice and is not a complete JLPT course: learners also need grammar, vocabulary, listening, and reading-comprehension work.